Hacktoberfest 2026: the issues maintainers tagged for October, open and beginner-friendly. Browse Hacktoberfest issues

ngit CI: test.yml lane is being dropped or phantom-failing at the coordinator level (operator investigation needed)

Open
  • Is something wrong?
#520 2 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

@Amperstrand is already working on this.

Since Oct 7, 2026.

  • #740 by @Amperstrand — open

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
35/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Active
Tech stack
go
Domain
ci-cd

Research direction

Start with the coordinator logs for the test.yml invocations on d622688b, fd582973, and 8e7a2e5c; compare them with the published 9841 jobs and 9842 conclusions described in the issue. Read .ngit/README.md for the runner and timeout constraints. Done means establishing whether these runs were dropped, died before publishing, or failed after jobs completed, and reporting the evidence.

Written by the indexing model from the issue text.

Description

Symptoms (relay evidence, kind 9841/9842/39842, coordinator 765cd47b…)

1. Dropped entirely. For d622688b (current main tip, pushed today) the test.yml lane produced zero events of any kind — no 9842 conclusion, no 9841 job results, no 39842 progress — while its three sibling lanes on the same push ran and concluded normally:

lane on d622688b 9842 9841 jobs
go-test.yml success go-checks: success
repro-check.yml failure (repo-side, see #519) binaries: failure
build-package-binaries.yml failure (repo-side, see #477) 3 jobs published
test.yml — none — — none —

2. Phantom red when it does run. On fd582973 and 8e7a2e5c the lane concluded failure while every published job was green — all module-matrix go-test legs, main-test, build-purity, contract-lint, deps-and-imports — with no __ngit_ci_timeout__ marker for these runs. A workflow that fails with no failing job means a job never started/published, or the invocation was killed between job completion and conclusion.

Historical note: one __ngit_ci_timeout__ 9841 exists (commit 15600ca7, 2026-09-21) — the 1800 s act ceiling — so both drop-modes and timeout-modes have now been observed for this lane.

Why this looks coordinator-side, not repo-side

  • The w tag content-sha of every executed workflow matches origin/main byte-for-byte — the runner is not executing stale files.
  • test.yml is by far the heaviest lane (16 jobs) on a runner documented at 4 vCPU / 4 GiB with NGIT_CI_MAX_CONCURRENT_JOBS=2 and a 30-minute per-invocation ceiling (.ngit/README.md).
  • The sibling lanes on the same pushes behave normally.

Ask

  1. Operator (DQ05): coordinator logs for the test.yml invocations on d622688b (never started?) and fd582973/8e7a2e5c (concluded failure with all-green jobs) — the relay cannot distinguish "dropped from queue" from "died before first publish".
  2. Repo-side mitigation (optional, if load is the cause): split the lane the way the release pipeline was sharded — e.g. module-matrix halves or moving main-test+contract-lint checks into go-test.yml's shape — so the heaviest invocation stops being the one the coordinator drops.

Until resolved, treat a green-or-missing test.yml as no signal: its absence on a push is not a pass.

Dominant language
Go
Stars
12
Forks
14
Avg merge
1d 6h
Merged PRs (30d)
217

Getting set up

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from OpenTollGate/tollgate-module-basic-go

All issues in OpenTollGate/tollgate-module-basic-go

Similar issues

More Go issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.